And that's a wrap from me (I'm about to run out of battery, but I've made it throgh the day) and what a fantastic day it's been - a great great day, I've really enjoyed it. I'm closing out the live blog now.

The closing remarks are optimistic, expectations for next year content indispensable to AI innovation in peer review research integrity a glimpse into the next chapter of this industry more new ideas new disrupters emerging some topics we will still be talking about

Close to closing remarks I really like Caroline's point that we are uniquely placed to help the creation of AI that is good for science. Stephen - huge potential commercial upside. Jess - new commercial models, but ones that we can communicate, and models that are flexible. Laura - the way content…

Caroline has a call to try to avoid fragmentation. She wants to identify the areas of collaboration. Laura - trust, integrity, TRACE, things we can't solve as individuals. Atul - simple standards. Barbara - making sure the content is discoverable - we make impact possible. Jess - the systems are aw…

So it's going to come down to culture. AWS were great at creating small pockets of innovation - even in a behemoth of an org.

Question from Stephen - will ai accentuate the scaling inequalities we have in the industry, or will it break them making it easier for small publishers to survive better? (IMO on the one hand AI gives the power and agility to small teams, but larger organisations can get to market at scale). Atul…

Do they think we will need new roles? Or same roles augmented. Basically - yes to both. That's the story on that question. That's it.

We need to give explicit permission to get people to take risks.

Question - is this a war time moment for these companies, or are they going to remain more collaborative - For Laura people are struggling with the idea that they can take more risk. Atul describers it as a fundamental workforce transformation process. Atul does not think we are in war time - it's…

Barbara mentions the growing voices around AI worries - (a good link on this is - this one - https://www.writingruxandrabio.com/p/curing-cancer-wont-redeem-ai). She asked an ethicist to come in and help them situate their concerns. He talked to them about purity and ethics - a really interesting pe…

Caroline asking about AI and what might be areas of disruption. Jess talks about the burden on peer review, and the increasing volume of requests coming in. So this is a core place. There is an opportunity to rethink roles, how might we make editor and publisher roles better. Laura is seeing the sa…

Caroline asks - who does not trust us - for me I think one thing that we might need to think about is that agents are not cognizient and so they need explicit signals, rather than implicit signals (I could be wrong about this because the LLM has enough world knowledge, but that world knowledge curr…

Ok, so we have an agreement that just is important, but given that no one is going to say that we need to be moving towards a low trust ecosystem, and a lower trust relationship with our communities, which means that there is a sense in which this question is potentially an empty question. It might…

On the question of operationalising trust: Stephen - for him it's about the process, and creating a high quality process. Barbara cites Susan Fisk - trust coming from a combination of warmth and competence, and the warmth is the related to the intent. For Jess it's an imperative as it is connected…

Well - last session - here we are!! Featuring: Atul Goel, President, KnowledgeWorks Global Ltd Laura Hassink, Managing Director/President for STM Journals, Elsevier/RELX Steven Heffner, Chief Publication Officer and Managing Director, IEEE Barbara Kline Pope, Executive Director, Johns Hopkins Unive…

I ask the panel about whether they are optimistic or pessimistic about the future - optimistic, optimistic, optimistic, realistic, realistic. Rob thinks that there is going to be some pain.

This was long but good. Now we have 15 mins for questions. I can't think of anything to ask.

Kim - thinks the big decision we need to make are submission fees - need to increase the friction to manage the incredible growth. Witold - thinks the most important is to train libraries so that they can train their community - PhDs who then train their professors, basically increase knowledge. Bu…

Taking now about what guidelines should be created Frontiers have created an AI playbook - https://www.frontiersin.org/ai-playbook/home - this is a living document - right now for peer review it says "Humans remain accountable: AI can support process but must not be the decision-maker or the source…

We now enter a bit of a cross discussion on the panel - the question on retraction, and being able to tell the story of the paper - comes up. There is agreement that this does not happen today but that there is value in giving these agents context.

Some discussions on what we might need to help with all this mess - and my message on this is that things are going to get a lot weirder before we can consolidate our ideas about the right future model.

The arXiv policy - https://blog.arxiv.org/2026/10/01/updated-rate-limit-policy/ - has come up. Kim points out that while the principle is interesting, it will likely be hard to enforce. Kim mentions submission fees. She would love quotas but she would love institutions to be more involved in these…

Jonathan - why not start with the idea that AI can do everything, and work backwards from there. Rather than the other way around. This is what you might do if you started as an AI native organisation. For internal facing things they have a high tolerance to blast radius as they have a small organi…

Oh, an interesting point - first time it's come up - and something we have talked a lot about in BMJ - AI can not only help against slop - we could - and we should - use these tools to make papers better - there are far more papers that AI could be used to help - and the upside is so much larger wh…

Kim points out that AI can help us with automated rejects, e.g. if there is a paper with a duplicate submission. That is not asking it to automate judgement, but automating where we have hard guidelines already. (EXACTLY!).

Jonathan - the systems their company is building is aiming to help publishers understand how these systems work.

Next Jonathan - who will be talking about agentic consumption infrastructure. all of their focus is on transparency in this context. Jonathan points out that the LLMs are narrating out what they are going to do (I heard that interacting with non RFLD's non aligned LLM is like speaking to a mad wizz…

Maria - humans need to be in control of final decisions. (I disagree with this, I don't think it needs to be applied in every case). So they are moving away from tick boxes - they want the handling editors to provide comments, not just box ticking. They are making some policy changes around things…

Kim - danger of race to the bottom - you don't want to get so fast that you - someone said you need to think of it like Guinness - worth eating for - and a great product - peer review times might not be the best thing to optimise for speed - more important to do it right than to do it fast. And mak…

Next up - a panel session - Kim Eggleton, Head of Peer Review & Research Integrity, IOP Publishing Maria Kowalczuk, Research Integrity Auditor Manager, Frontiers Jonathan Woahn, Co-founder, Cashmere

They are now trying to build a tool to support teachers for creating e-learning materials - so they are very forward thinking in their adoption of the tools. They need cooperation with publishers to move this forward - so they need clear policies on what use is allowed (not the first time this ques…

OK OK, so - this topic is directly in line with a lot of the topics that I have been doing a lot of thinking about. (When Witold is speaking about policy development and the desire for transparency - for me it's important that our policies are directive in terms of obligations - but I think should…

Questions from the librarian lens - speed vs integrity, scale vs accountability, transparency vs complexity, governance vs innovation. (IMO - These are reasonable questions, but the technology does not pause to take these into account, and nor does human behaviour).

(We love you Witold!)

Next up - Witold

As a peer reviewer Indira - Overall her perspectives tell us that these tools are having an impact on her work, but her account so far is primarily grounded in just her own experience, posting these questions - but not resolving the issues, or offering a path. She mentions AI reviewers as a strateg…

(this is the paper the Rob mentioned - https://arxiv.org/abs/2602.20946 Some Simple Economics of AGI)

We have a video feed from Indira - giving her perspective as a researcher, editor, peer reviewer. AI is impacting all of these areas of her life. She studies the LLMs as an object of study. She also uses them as tools. It's clear they are having a key impact on her research practice. She reflects o…

this post - https://blog.mulvany.net/posts/dark-knowledge

Rob is talking about the binding conturing on growth no longer being intelligence, but rather the ability to verify (however what needs to be take into account is - for me - Will we end up in a hollow economy - or an augmented economy. (I think I coined the term Hollow Papers earlier this year)

And we are back - big session on AI an integrity coming up. Trust by Design: Who decides where AI stops, and humans begin?Moderated by Rob Johnson, Managing Director, Research ConsultingWith Kim Eggleton, Head of Peer Review & Research Integrity, IOP PublishingMaria Kowalczuk, Research Integrit…

Question asking if there is a simple definition of a "domestic journal". The answer is whether the journal has a cn number, rather than an ISSN (harder to get). (what are the requirements to get one??). Ownership is important too - must have Chinese ownership, e.g. Wiley can help co-publish these t…

Question now about retractions - person asking the question mentioned that the speakers didn't talk about this, but rightly one of the speakers points out that China policy focussing on integrity issues. Those standards include looking at retraction rates of journals. Of course.

A question from the audience on the difference between English Language and Chinese language journals. The answer is a little opaque but it seems like the key is that their evaluation systems are applied equally to English language and Chinese language journals - so it seems to me from the answer i…

Quality comes up time and again, but we don't have total clarity yet on what that means, and we also need to start thinking about things like the potential impact of some of the things we were discussing earlier in the conference, like the TRACE initiative.

We are in QandA now.

Bob Howard, Executive Vice President, Research, SAGE Publishing Bob points out that there has been a lot of changes in how china is evolved. Things like governance and pushing down on paper miills. The government no longer making cash payments for publishing in high IF journals (though some payment…

Dr Sun Yu - talking about the excellent action plan This is a 5yr program with funding of about $35M. So this is the pot that goes into part of their journals strategy. A bit problems for China is that while their output has grown the contribution to China specific journals has actually shrunk. Thi…

This session: From output to influence: China, and the future of scholarly communication With Yu Sun, Associate Editor-In-Chief, Tsinghua University Press,Ting Yue, OAJ team lead, National Science Library, Chinese Academy of Sciences, and Bob Howard, Executive Vice President, Research, SAGE Publish…

And we are back from lunch(Over lunch I had a lovely chat with Patrick Starke - https://www.linkedin.com/in/patrick-starke-712410b3 - CEO of https://imagetwin.ai/ - of note he says that AI can now generate images that are indistinguishable to the human eye, and if we run images through a process th…